W&M ScholarWorks W&M ScholarWorks

Dissertations, Theses, and Masters Projects Theses, Dissertations, & Master Projects

1979

The spatial distribution and agonistic behavior of an eastern gray The spatial distribution and agonistic behavior of an eastern gray

squirrel (Sciurus carolinensis) population on Jamestown Island, squirrel (Sciurus carolinensis) population on Jamestown Island,

Virginia Virginia

Marston Earl Youngblood College of William & Mary - Arts & Sciences

Follow this and additional works at: https://scholarworks.wm.edu/etd

Part of the Ecology and Evolutionary Biology Commons, and the Zoology Commons

Recommended Citation Recommended Citation Youngblood, Marston Earl, "The spatial distribution and agonistic behavior of an eastern gray squirrel (Sciurus carolinensis) population on Jamestown Island, Virginia" (1979). Dissertations, Theses, and Masters Projects. Paper 1539625053. https://dx.doi.org/doi:10.21220/s2-7kds-c051

This Thesis is brought to you for free and open access by the Theses, Dissertations, & Master Projects at W&M ScholarWorks. It has been accepted for inclusion in Dissertations, Theses, and Masters Projects by an authorized administrator of W&M ScholarWorks. For more information, please contact scholarworks@wm.edu.

THE SPATIAL DISTRIBUTION AND AGONISTIC BEHAVIOR

OF AN EASTERN GRAY SQUIRREL

(SCIURUS CAROLINENSIS) POPULATION ON

JAMESTOWN ISLAND, VIRGINIA

A Thesis

Presented to

The Faculty of the Department of Biology

The College of William and Mary in Virginia

In Partial Fulfillment

Of the Requirements for the Degree of

Master of Arts

by

Marston E. Youngblood, Jr.

1979

APPROVAL SHEET

This thesis is submitted in partial fulfillment of

the requirements for the degree of

Master of Arts

Aut

Approved, December 1979

C. Richard Terman, Ph.D.

Stewart A. Ware, Ph.D.D

nnaoCcW Q q. aMitchell A.Byrd, Ph.D.'

prman J. Fashing

TABLE OF CONTENTS

ACKNOWLEDGMENTS . . .

LIST OF TABLES..................

LIST OF FIGURES ..................

ABSTRACT. . . . . . . s = . . . .

INTRODUCTION......................

REVIEW OF LITERATURE. . . . . . .

STUDY AREA.........................

MATERIALS AND METHODS ...........

Trapping Procedures...........

Observational Procedures . . .

Removal and Release Procedures

Statistical Procedures . . . .

RESULTS ...........................

Population Description . . . .

Spatial Distribution .........

Removal and Release...........

Behavioral Observations. . . .

DISCUSSION.........................

APPENDIX...........................

LITERATURE CITED..................

VITA................................

ACKNOWLEDGMENTS

The author expresses his appreciation to Mr. Glen Bean who patiently

and skillfully constructed the live-traps for this study. Ms. Sarah Gay

Wilkins is also deeply appreciated for her help in illustration of the

figures, as is Mrs. Jean Belvin for final preparation of the manuscript.

The author acknowledges Mr. William B. Sullivan, Superintendent, and

Mr. Ralph Maxwell, Chief Ranger, of Colonial National Historical Park

and their staff for their professional assistance in conducting this

study on Jamestown Island.

Dr. C. Richard Terman is thanked for his long-standing encouragement

and advice to the author during the course of this research project.

The author is also grateful for the critical comments and reading of the

manuscript by Drs. S. Ware, N. Fashing, and M. A. Byrd.

Finally, the author is indebted to his parents for their support and

encouragement during this period of study.

LIST OF TABLES

Table Page

1. Sex and Age of Gray Squirrels Captured by LiveTrapping...................................... . 19

2. Frequency of Captures of Squirrels in Live-traps.............. 20

3. Trap Response of Gray Squirrels to Urine ApplicationExperiment.......................................... „ . 23

4. Home Range Areas of Resident Animals............................38

5. Average Per Cent Overlap of Squirrel Home Ranges.............. 41

6. Average Overlap Indices of Squirrel Home Ranges ........... 43

7. Average Distance between Activity Centers of NearestSame Sexed Animals . 45

8. Average Distance between Activity Centers of NearestOpposite Sexed Animals.................................. 48

9. Comparisons of Average Shift in Activity Centers.............. 50

10. Comparisons of Dominance Index by Sex and Season.............. 54

11. Spearman Rank Correlations of Dominance Index .............. 55

LIST OF FIGURES

Figure Page

1. Map of Study Area.................................... 9

2. Distribution of Overstory and Major UnderstoryVegetation in Study A r e a ................. . 1 0

3. Urine Treatment Experimental Trap Locations. . . . 1 2

4. Capture Effort by Month......................................... 21

5. Male Reproductive Condition during 1975 and 1976 ......... 24

6. Female Reproductive Condition 1975 ......................... 26

7. Female Reproductive Condition 19 76 . . - . . 27

8. Reproductive Rate of Adult Females . . . 2 8

9. Average Male Body Weights........... . . 29

10. Average Female Body Weights.....................................30

11. Home Range Ellipses for Residents during Spring 1975 . . . 33

12. Home Range Ellipses for Residents during Summer 1975 . . . 34

13. Home Range Ellipses for Residents during Fall/Winter1975 .................................................. 35

14. Home Range Ellipses for Residents during Spring 1976 . . . 36

15. Home Range Ellipses for Residents during Summer 1976 . . . 37

16. Dominance Index Scores for Males and Females .............. 52

17. Levels of Aggressive Interactions during Summer 1975and 1976 ............................................. 56

18. Levels of Aggressive Interactions Adjusted forPopulation Density during 1975 and 1976........... 58

19. Orientation of Aggressive Interactions during Summerof 1975 and 1976 .................................... 59

vi

ABSTRACT

The purposes of this study were to first, apply a bivariate proba­bilistic home range model (Koeppl e_t al_. 1975) to examine the spatial distribution of a natural gray squirrel population. The second aspect of this study was to examine the relations between agonistic behavior of gray squirrels with their spatial distribution. The approach- avoidance behavior of gray squirrels in response to the stimulii of urine scent marks are tested as a final aspect of this study.

Livetrapping and direct observations of squirrels were conducted for 18 months, resulting in a total capture of 37 males and 23 females which were marked and released. Gray squirrels were captured an average of 4.5 + 0.756 times (Mean + SE). Frequency of captures ranged from 1 to 29; approximately 35% of the animals were captured once, 60% were captured from 2-10 times, and 5% were captured more than 25 times. The results of the urine application experiment did not indicate a signifi­cant approach or avoidance response of gray squirrels towards entering urine treated live-traps. Furthermore, no scent marking was observed.

The data indicate that males on the average have larger home ranges (2.28 ha) than females (1.92 ha), but the difference is not significant. No clear age related difference in home range size is indicated. How­ever, subadult squirrels were more mobile and had larger home range areas than adults (2.5 and 3.12 ha versus 2.34 and 1.6 ha for both males and females). Seasonal changes in home range size were not demon­strated in this study.

Significant differences in spatial distribution of the study popu­lation were indicated by changes in the percentage of home range overlap and overlap index. These differences can be attributed to both sex related and seasonal effects of the spring-summer breeding period and fall dispersal. The significant age related difference in distance to the activity center of nearest same sexed animal was linked with the increased distance between the activity centers of subadult and other males in the spring and fall/winter seasons. Avoidance by subadult males of the more aggressive and dominant adults during these periods could account for the significant age related differences.

Established animals in the Jamestown study population had stable home ranges for the duration of this study. The removal/release pro­cedure results of this study confirm that resident gray squirrels are likely to return when released within 1 km of their home woodlot.

Adult males in this study tended to have higher Dominance Index (D.I.) scores than females, but this difference was significant only for the spring observational periods. Body weight of the animal was positively correlated with home range overlap. The data suggest that subordinate gray squirrels were avoiding contact with dominant con- specifics in the study area. Dominance was not associated with actual estimated size of home range.

vii

The results that dominance was not shown to be associated with home range size as by Pack et al. (1967), and that residents were not con­sistently dominant over immigrants as shown by Thompson (1978) may be attributable to use of a feeding station in this study, since noticeable increases in the levels of aggressive interactions began upon its establishment on the study area. Utilization of the home range model of Koeppl e_t aJ. (1975) resulted in values for gray squirrel home ranges within the limits of areas reported in the literature and provided a quantifiable basis for analyzing the spatial distribution of the popu­lation .

viii

SPATIAL DISTRIBUTION AND AGONISTIC BEHAVIOR

OP AN EASTERN GRAY SQUIRREL POPULATION

INTRODUCTION

The eastern gray squirrel, Sciurus carolinensis, is one of the most

intensively studied small mammals in reference to home range (Flyger

1960; Taylor 1966; Doebel 1967; Donohoe and Beal 1972; Cordes and

Barkalow 1972; Doebel and McGinnes 1974; Thompson 1978) and behavior

(Bakken 1959; Sharp 1959; Taylor 1966; Pack et_ al . 1967; Horwich 1968;

Brady 1972; Cordes and Barkalow 1972; Bland 1977; Thompson 1977a, 1977b,

1978). Studies evaluating dominance, dispersal, and social heirarchy

indicate gray squirrel woodlot populations have a relatively stable

social system, with an individual squirrel’s position in the social

heirarchy shown to be correlated with age, sex, and body size, but not

conclusively associated with home range size (Flyger 1960; Cordes 1965;

Taylor 1966; Pack et al. 1967; Thompson 1978).

D. C. Thompson (1978: 326) succinctly summarized the social system

of the gray squirrel as determined by his research:

The area used by males and females expands after weaning, then stabilizes and remains the same in location and extent for life. The home range of an established indi­vidual is broadly overlapped by the home ranges of several other animals. Each established individual is regularly in contact with only a limited number of recognized neighbors with which it has well-established dominance relationships. Individual recognition promotes lowered aggressive levels between neighbors which allows each squirrel to use its entire home range evenly. Aggressive behavior is directed toward strange squirrels, either young or immigrants, which attempt to enter this system.Thus, the established individuals hinder the settlement of new animals. Young squirrels born in a given locality have a greater chance of establishing than do immigrants.

Thompson’s (1978) definition of the gray squirrel social system parallels

the results of Bronson (1964) with another sciurid, Marmota monax. The

3

woodchucks that he observed appeared to be organized into a complex of

dominance-subordination relationships which were stable regardless of

the location of an interaction within the system of overlapping home

ranges.

Few recent investigators of squirrel home range and behavior have

utilized descriptive probabilistic models to estimate the area used by

an individual animal. The method most often used for determining

squirrel home ranges is the minimal polygon method (Flyger 1960, Doebel

1967, Taylor 1966, Pack et al. 1967, Cordes and Barkalow 1972, Thompson

1978). Adams (1976), however, used the bivariate normal model of

Jennrich and Turner (1969) to generate probability ellipses to estimate

fox squirrel (Sciurus niger) home ranges. No researchers of squirrel

behavior to date have employed the probabilistic home range model of

Koeppl et a2L. (1975) to investigate the spatial distribution of a gray

squirrel population. Consequently, one of the purposes of this study

is to apply the bivariate, probabilistic home range model of Koeppl e_t

al. (1975) to examine the spatial distribution of a natural gray

squirrel population on Jamestown Island, Virginia.

The advantages that exist in studying a protected (from man) popu­

lation of gray squirrels on Jamestown Island are twofold. First, it is

an opportunity to study a well defined, isolated, population of a

readily observable small mammal. Second, a field study presents the

opportunity to observe behavior of the gray squirrel in the wild with

the influences of weather, predation, and a natural habitat contributing

to behavioral interaction. These important environmental factors would

otherwise be lost under laboratory or semi-natural conditions.

Watson and Moss (1970) observed that behavior involving the

dominance of animals over others often operates via aggression to effect

patterns of spacing. In the home range system, as opposed to a

territorial system, a species is attached to a fixed area of land but

does not defend it; instead, the animal's portable personal sphere, the

area around a solitary individual, is defended against intrusion

(McBride 1971). Agonistic interactions are concerned with expelling

the opposing conspecific from this area or forcing him to submit. In

either case, a dominance-subordination relationship is established,

thereafter expressed spatially (McBride 1971) . The second aspect of

this study is to examine the relations between agonistic behavior of

gray squirrels with their spatial distribution.

There are reports of gray squirrels using scent marking in woodlot

areas (Taylor 1966, 1968; Barkalow and Shorten 1973). Taylor (1966,

1968) hypothesized that marking points function as an act of communi­

cation among a population of squirrels. If scent marking is interpreted

as a form of social behavior in the gray squirrel, then it may be a

factor in individual recognition between neighboring animals and thus

influence the spatial distribution of the population. The approach-

avoidance behavior of gray squirrels in response to the stimuii of

urine scent marks will be tested as a final aspect of this study.

REVIEW OF LITERATURE

W. H. Burt (1943) provided an early operational definition of home

range. He restricted the home range concept to "that area traversed by

the individual in its normal activities of food gathering, mating, and

caring for young." Furthermore, he suggested that once an individual

established a home range it normally remained in that area for life,

with the exception of occasional exploratory movements into areas which

should not be considered part of the home range.

The size of a home range may vary with sex, age, and season; and

home ranges of different individuals may even overlap to varying

degrees. One of the simplest methods of estimating the home range of

small mammals is derived from drawing the smallest polygon which

contains all the locational data points for an individual and measuring

the resulting area. The problems that arise from this minimum polygon

method are that the shape of the polygon and its area depend on the

order in which the locational points are connected, and the areas of

minimum polygons have a tendency to increase in direct proportion to

the number of locational fixes in the sample (Jennrich and Turner 1969) .

A circular home range model was proposed by J . C. Calhoun and

J. U. Casby (1958) based on the mean recapture radius of small mammals.

The capture radius (r^) for a particular capture point (pi = x^, y- ) was

defined "as the distance from p^ to the geometric center (p = x, y) of

the capture points" (Jennrich and Turner 1969: 229). A value was

computed from the recapture radius and estimated home range size was

determined for the defined circular area. The drawbacks to a circular

home range model are that it underestimates size in comparison to long

narrow ranges covering the same area, and it is based on a strict

assumption of circular symmetry.

Jennrich and Turner (1969) note that many non-circular home ranges

have been reported for mammals, and propose a model that could measure

circular as well as non-circular home ranges. They assume that the

intensity with which an animal uses its habitat is expressed by a

bivariate normal distribution. In addition, this utilization distri­

bution is best defined by concentric constant density ellipses. Home

range for their elliptic model is defined as the "area of the smallest

region which accounts for 95% of an animal's utilization of its habitat"

(Jennrich and Turner 1969: 233). They recommend their estimate of

home range because of its lack of sample size bias, the non-assumption

of circularity, and an independence from orientation of home range.

A contrasting approach is presented by Van Winkle (19 75) in which

he discusses the elliptical home range model offered by Mazurkiewicz

(1969, 1971) that uses three sample statistics to provide information on

orientation of home range.

These three statistics are the lengths of the major and minor axes (principal components) of the ellipses of concentration and the angle of inclination of the major axis with respect to the original coordinate system. Mazurkiewicz (1971) assumed that the angle of inclination was an index of direction of movement preferred by an individual and that the ratio of the principal axes was an index of the degree of preference for this direction.In her studies on the home ranges of the redback vole, Clithrionomys glareolus, no circular areas were found; in all cases the home range had the shape of a flattened ellipse, with a mean ratio of 3:1.

The bivariate normal distribution without the assumption of circularity

of an individual's utilization distribution provides a general and

flexible probabilistic home range model that is adequate for depicting

the home ranges of many animals inhabiting homogeneous habitats

(Van Winkle 19 75).

Another probabilistic bivariate home range model is presented by

Koeppl et a l . (1975). They propose a general model for inter- and

intraspecific comparisons as well as inferences concerning internal

structure of the home range. For this model, the axes of the ellipse

are weighted according to the frequency of occurrence of the individual

at a particular location. Biases due to orientation of locational

data and sample size are corrected by the use of eigenvalues of the

variance-covariance matrix derived from the data coordinates and by

incorporating the F-statistic in calculation of the ellipse (Koeppl et

al. 1975). Thus, their home range model is a confidence ellipse which

is utilized in determining the probability of finding an individual at

a particular location.

Hayne (1949) defined a point which is the geographic center of all

points of capture (or observation) of an individual as the center of

activity for that animal. Koeppl ej: .al. (1975) believe that the

activity center is more than a geometric center of a scatter of points.

They postulate the concept of a center of familiarity, a location in an

individual’s home range in which it is most familiar and secure. The

advantages of their model are that the estimated areas, A p , reflects

the confidence in home range size and location, given a finite number

of home range coordinates, and that it permits inferences concerning

an animal's relative familiarity with any point within its home range

(Koeppl ej cfL. 1975) .

STUDY AREA

The study site is approximately 2.4 hectares (6.01 acres) in size,

and is located on Jamestown Island, Virginia, which has been under

protection of the National Park Service since 1934 as part of Colonial

National Historical Park. The study area is situated at the eastern

edge of the New Town meadow, and is bounded on the other three sides

by the James River, Orchard Run, and the Jamestown Island loop road

(see Figure 1). The geologic characteristics of the area are a basal

layer of sand which is topped by a few feet of clay and silt merging

into topsoil, with a maximum elevation of less than 10 feet above mean

low tide (Cotter 1958).

Characteristic tree species in the overstory of the study area are

loblolly pine (Pinus taeda), pecan (Carya illinoensis), black cherry

(Prunus serotina) , tulip popular (Liriodendron tulipfera), several

different oaks (Quercus spp.), and bald-cypress (Taxodium distichum).

Species common to the understory are eastern redcedar (Juniperus

virginiana) , American holly (Ilex opaca), dogwood (Cornus sp.), wax

myrtle (Myrica cerifera) , and wild grape (Vitis sp.). Various grasses,

sedges, poison ivy (Rhus radicans), and crownbeard (Verbesina

occidentalis) form the ground cover of the area (Figure 2).

Figure 1. Map of study area.

EW TOWN M EAD O W PITCH

& TAR SWAMP

FSTS

ISLAND " LOOP

ROAD

OB

ORCHARDRUN

▲ FS — Feeding Station ■ OB — Observation Blind

♦ RP - Release Point• TS Sta

24 m

RP

VAC'""-’

10

Figure 2. Distribution of Overstory and Major Understory Vegetation in Study Area.

(Encircled symbols on map indicate high density of a species.)

• Trap S t a t i o n

Pt - Pinus ta eda C i - C a r y a i l l i n o e n s is Q - Quercus spp.Ps- Prunus s e r o t in a J v - J u n i p e r u s v i r g i n i a n a T d - T a x o d i u m d i s t i c h u m Lt - L i r i o d e n d r o n t u l i p i f e r a lo - I lex op a c a

M c - M y r i c a c e r i f e r a

24m

MATERIALS AND METHODS

In March, 19 75, a trapping grid was established on the study area.

Grid lines were laid on a 5 column (23° West of North) by 7 row (67°

East of North) design with 20 m inter-station spacing. The lines of

the grid were shortened as necessary to conform to the natural boundaries

of the study area. The result was an irregular shape for the overall

grid layout (Fig. 3) of 23 trap stations. Live traps were placed two

per station within 1 m of the grid stake.

Each live trap was 25 cm X 25 cm X 61 cm rectangular pine box with

hardware cloth at one end for ventilation and a gravity fall trap door

at the opposite end (after Mosby 1969). The hardware cloth reinforced

trap door was released when the investigating animal tripped a baited

treadle. Shelled feed corn, D&G lab chow, raw peanuts in the shell,

and whole pecans served as bait during the course of the study.

Trapping Procedure

Fourteen-day trapping periods were conducted each month from

April 5, 1975, until August 14, 1975. From September 1, 1975 to

September 12, 1976, a continuous trapping schedule of 7 days set traps

followed by 7 days closed traps was followed. Because of the bimodal

nature of the gray squirrel daily activity period (Horwich 1968),

trapping was conducted during either early morning or late afternoon

hours. In the morning, traps were opened and set shortly after dawn,

and were checked and closed 5 to 8 hours later. On days in which

trapping was done in the afternoon, traps were opened and set at

midday and checked for captured animals at sundown.11

12

Figure 3. Urine treatment experimental trap locations.

• M• M

• M• M

• M

• M• M

• M

• M• M

• M

M : M ALE URINE M A R K E D TRAP

F : FEMALE U R IN E M A R K E D TRAP

13

In March, 1975, 7 adult and subadult gray squirrels were live-

trapped at a private home in Williamsburg, Virginia, located approxi­

mately 5 km from the Jamestown Island study area. These gray squirrels

(4 females and 3 males) were housed at the Laboratory of Endocrinology

and Population Ecology of the College of William and Mary, and supplied

with food (D&G Laboratory Chow) and water ad libitum. From March 22,

1976, to August 25, 1976, 6 drops of urine collected from these

squirrels or 6 drops of distilled water was introduced into selected

traps at the beginning of each trap day in addition to the normal bait

allocation. The urine was collected by placing screen-covered aluminum

foil trays underneath the wire mesh floor of solitary, caged gray

squirrels and then pipetting deposited urine into clean 10 ml glass

vials for storage. The urine was kept frozen until required for appli­

cation to the live traps. Only urine from adult males and females was

used, and urine from each sex was collected and stored separately.

Application of urine or water to the live traps was performed

systematically. At each station, one trap was treated with water and

the other with urine (either male or female). Six drops of the

appropriate urine or distilled water was applied with a dropping pipette

to the treadle of each trap as it was set for the days trapping. The

total number of water controls was 2 3, while there were 12 male urine

and 11 female urine treated traps (Fig. 3). The schedule of trapping

remained unchanged (7 days set followed by 7 days closed). Traps were

consistently marked with the same scent treatment from March 22, 1976,

to August 25, 1976. Number of captures for each trap per station were

recorded at the conclusion of each trap day.

14

Captive gray squirrels were removed from the live traps by placing

a burlap sack over the entrance of the trap, releasing the door, and

stimulating the animal to leave the trap and enter the sack by blowing

through the hardware cloth end of the trap. For each captured gray

squirrel, certain basic information was recorded. Squirrels were sexed,

weighed with a Pesola 1000 g scale (after June 1975) , and aged by

pelage characteristics (Sharp 1958; Barrier and Barkalow 1967). Repro­

ductive condition of males was assessed by noting testes descent via

palpation (either testes scrotal or testes non-scrotal condition

recorded) and by observing scrotal characteristics such as general size,

pelage, and coloration (Pudney 19 76). Mature male squirrels typically

have large testes contained in well formed scrotal pouches which are

usually bald and heavily pigmented (Pudney 1976). Reproductive infor­

mation recorded for females included condition of the vagina (perforate

or imperforate), size and pigmentation of nipples, and the presence of

pregnancy or lactation as determined by palpation (Nixon and McClain

1975) . Locational data (grid coordinates) for each capture were also

recorded as well as general weather conditions.

Three marking techniques were employed to facilitate identification

of individuals. Gray Squirrels were toe clipped according to a

numerical code (Taber and Cowan 1971). Secondly, an ear punch was used

to perforate each ear before inserting a Monel 5/16M ear tag (National

Band & Tag) backed by a Monel washer and a colored celluloid 3/4"

washer which was crimped into place. The third marking technique

consisted of daubing geometric designs upon the pelage of the gray

squirrel with Nyanzol D (Nyanza Inc.). This dye is a permanent, non­

toxic, black dye frequently used for marking animals. Dye markings were

15

lost as the gray squirrels shed their pelage; therefore, animals were

re-marked often upon recapture to assure permanence and distinctiveness

of the dye pattern. After completion of marking, examination, and

physical data collection, each squirrel was released at the point of

capture.

Observational Procedure

From April, 1975, until September, 1976, detailed observations and

records were made on individual gray squirrel movements and behavior.

Movements of animals over the study area were recorded on maps during

each period of observation. Using descriptions of typical gray squirrel

behavior (Bakken 1959; Sharp 1959; Taylor 1966; Pack et al. 1967), an

ethogram of agonistic, investigative, sexual, and feeding behaviors was

devised (see Appendix). The time, location, and type of behavior

observed for each squirrel during the course of each observational period

was logged on a recording sheet. Interactions between squirrels were

recorded sequentially until the animals were lost from sight. Weather

conditions for each observation period were also recorded.

Observation times were mainly confined to the principal morning and

late afternoon activity periods of the gray squirrels (Bakken 1959;

Horwich 1968). Morning observations were made between 06:30 and 10:30

for a period of 3 to 4 hours. Late afternoon observations were made

between 16:00 and 20:30 for corresponding lengths of time. Gray

squirrels were initially observed by slowly walking through the study

area and pausing for 30 min. to an hr. as active squirrels were located

before moving on to a different portion of the area. In June, 1975, an

observation blind was placed at a central point in the study area

16

(Fig. 1). From June, 1975, until September, 1976, most observations of

gray squirrel movements and behavior were made from the blind. Good

visability of the major portion of the study area was possible due to

the lack of obscuring ground cover during most seasons.

On June 21, 1976, a feeding station was placed at grid coordinate

C4 (Fig. 1) approximately 12 m from the observation blind. Pack et al.

(1967) have demonstrated that a feeding station will act as a focal

point to attract gray squirrels to an area, and also serve to stimulate

behavioral interactions between animals. The station was supplied

daily with 15-20 pecans at the beginning of each observation period.

The feeding station was discontinued on August 22, 1976, when suitable

bait (pecans) became unavailable.

Removal and Release Procedure

From August 15, 1976, to August 20, 1976, 6 resident gray squirrels

were removed from the study area and caged at the Laboratory of

Endocrinology and Population Ecology. Residents for the purposes of

this study were defined as animals that had been captured 5 times and/or

observed on the study area for at least one month.

On August 20, 1976, the 7 captive gray squirrels from Williamsburg

which had earlier been toe clipped and ear tagged for identification

were transported from the laboratory to Jamestown Island. At 08:03

they were released at a point 488 m from the Island loop roadway

boundary of the study area (Fig. 1). The live traps on the study area

were then opened and set, and the observation blind occupied following

standard procedures as described.

17

The resident Jamestown gray squirrels were held in the laboratory

for at least one week from date of capture, and on August 29, 1976, at

07:30 transported back to Jamestown Island. They were released from the

same point on the island at which the alien, Williamsburg gray squirrels

had previously been released. Throughout the periods of resident

removal, alien gray squirrel release, and then resident gray squirrel

release, the standard trapping procedures and schedule were employed.

Statistical Procedures

Extensive use was made of packaged statistical computer programs at

the College of William and Mary Computer Center to analyze the trapping

and observational data. Biometry programs developed by F. J. Rohlf

(Sokal and Rohlf 1969) were used to calculate the 95% confidence

ellipses for home range estimations (Koeppl et al. 1975) . Mann-Whitney

U tests, Spearman’s r, three-way ANOVA's, and Duncan’s range tests were

performed using the Statistical Package for the Social Sciences (SPSS)

versions 7.0 and 8.0 (Nie et al. 1975; Nie and Hull 1977). Occasionally,

statistics were manually computed using methods described in Sokal and

Rohlf (1969) and Siegel (1956). The level of statistical significance

employed was 0.05 unless otherwise noted.

RESULTS

Population Description

The results of 189 days of live-trap operation for the 18 month

duration of the study yielded a total of 60 captured gray squirrels

(Table 1). Of the total number of captures, 37 males and 23 females

were marked and released. Adult animals comprised 65% (39) of the

captures, while subadult and juvenile animals represented the

remaining 22% (13) and 13% (8) of captures. A chi-squire test of

independence of sex and age upon trap response was not significant 2(X = 3.83, 0.50 > p > 0.10), indicating no differential captures due

to sex or age.

Gray squirrels were captured an average of 4.5 + 0.756 times

(mean + std. error, Table 2). Frequency of capture ranged from 1 to

29. Approximately 35% (21) of the animals were captured once and 60%

(36) were captured from 2 to 10 times. The remaining three individuals

were captured more than 25 times and could be categorized as trap-

addicted .

Figure 4 illustrates trends of capture effort for the duration of

the study. Starting from April, 1975, the rate of 1.7 captures/day

increased to the June and August, 1975, peak capture rates of 4.2 and

4.3/day. Rate of captures were lowest in September, 1975, as shown

by the rate of 0.8 captures/day. The rate of captures irregularly

increased during the winter to a high point in February, 1976, of

1.9/day. A decline in captures occurred in the spring, culminating

in a rate of 0.0/day in May, 1976. A steady increase thereafter led

18

19

Table 1. Sex and age of gray squirrels captured by live-trapping. Chi-square test of independence of sex and age not signifi­cant, x^ = 3.83, 0.10<p<0.50.

Age MaleSex

Female Total %

Adult 27 12 39 65.0

Sub adult 5 8 13 21.7

Juvenile 5 3 8 13.3

Total

%37 23

61.7 38.3

60

100

20

Table 2. Frequency of captures of squirrels in live-traps.- Mean number of captures = 4.5, standard error = 0.756, standard deviation = 5.86, n= 60.

Times Captured Number of Animals

1 21

2 9

3 8

4 3

5 3

6 3

7 3

8 3

9 2

10 2

25 1

27 1

29 1

21

Figure 4. Capture effort by month.

I d 3 S

onv

'AVW

Nvr

A O N

I DO

o n v

in r

Nnr

AVW

COin CN o

AVQ d 3 d S 3 d n i d V D

1975

M

ONT

HS

1976

22

to another summer peak in July, 1976, which then dropped off to a

final low of 1.0/day at the conclusion of the study in September, 1976.

During the experimental urine application phase of live-trapping,

97 captures were recorded (Table 3). Of these captures, 43 animals

were caught in the control (water marked traps) while 34 animals were

captured in the male urine treatment and 20 animals in the female

urine treatment. A chi-square test of independence of urine treatment

and subsequent captures was not significant (X = 1.378, 0.50 < p <

0.90). Application of the urine treatment to live-traps did not

significantly alter the trap response of gray squirrels in the study

population.

The size of the Jamestown study population was estimated using

the live-trapping data for the peak summer capture periods of 19 75 and

1976. The technique employed was the Maximum Likelihood Estimate (MLE)

developed and tested by Nixon, et_ aT. (1967) for estimation of squirrel

abundance from live-trapping data. The equation used to determine the~ _ vnMLE is derived from the well known Lincoln index: ^ - —_ ’

1- (Enx/fxnx)where N = the estimate of the population, Znx = the total number of

animals handled, and Exnx = the total number of captures (Nixon et al.

1967). The MLE for July and August, 1975 was 38.1 animals. The same

period in 1976 produced a MLE of 15.2 animals.

Reproductive condition of captured male squirrels as assessed by

scrotal sac characteristics is shown in Figure 5. The largest number

of males with scrotal testes were captured in August, 1975. Relatively

high proportions of non-scrotal testes males were captured in July,

October, and December, 1975. The number of immature males with non-

scrotal testes remained high through January and February, 1976,

23

Table 3. Trap response of gray squirrels to urine application experiment. Chi- square test of independence of treatment and captures not signifi­cant, = 1.378, 0.50<p<0.90.

TreatmentCaptures

Males Females Total

Water 19 24 43

Male Urine 17 17 34

female Urine 12 8 20

Total 48 49 97

24

Figure 5. Male reproductive condition during 1975 and 1976.

^ W PX ^ OS 3 1 V M JO SH 38N 0N

29682719017595

25

before declining. Males with scrotal testes began increasing in March

and April, 19 76, with a peak in male sexual development occurring in

August-September 1976.

The reproductive condition of females is shown in Figure 6 for

1975 and Figure 7 for 1976. April and August, 1975, showed the

highest numbers of pregnant or lactating females in the population.

High numbers of females with open vaginae were captured in June,

August, and December, 1975. Females having closed vaginae were only

captured in April and May, 1975. In 1976, the highest number of

pregnant or lactating females were captured in April. A few imperforate

females were captured in the months of February and April, 1976.

Mature, perforate, females were captured more frequently in June and

August, 1976.

The reproductive rate for female squirrels was computed from the

capture data and is detailed in Figure 8. Reproductive rate is defined

here as the number of pregnant or lactating females divided by the

total number of mature females. Definite spring and summer peaks of

reproductive rate are evident in the study population of gray squirrels.

Intermediate levels of reproductive activity (0.50 - 0.55) occurred in

April and August, 1975, and July, 1976. A sustained high reproductive

rate of 1.00 occurred from March to May, 1976.

Average total body weights for captured males is plotted in

Figure 9, and for females in Figure 10. The body weight data (Figs.

9 and 10) must be presented with reservation due to the degree of

uncertainty arising from small sample sizes with high individual

variation, and the fact that the animals may have lost varying amounts

of weight while in the live-traps. Adult males showed a trend for a

26

Figure 6. Female reproductive condition 1975.

Q

E! ■! ■ 5 S 6 5 8 S

>-■a:

| 1 1 1 1 1 1 1-------co r— to m «*r co c i t— o

S31VW33 30 SH3flWflN

27

Figure 7. Female reproductive condition 1976

IMP

ER

FOR

ATE

Q.LUUi

UJI—<OCo

K

oc o° 5

< uz <CD -iLUOCCL.

O3<

z3

>■<

fidQ.<

fid<2

K K * ?z<

S31 V W 3 d dO S 3 3 S W n N

(9/61)

SH

iNO

W

28

Figure 8 Reproductive rate of adult females(numbers in parentheses are totalnumber of animals).

id 3 Sm o n v

i n rCN

COAVW

dVWcs

CN AON

CO 1 DO

I d 3SCN

o n v

i n r

00

CO AVW

oNO00 lOo

31VH 3 A l l3 n a O d d 3 d

1 9 75

MO

NTH

S 1976

29

Figure 9 Average male body weights(numbers in parentheses arenumber of animals).

28 B'O

CN

COI—_JZDa<

CO I— _I=3Q<COZDCO

CO111

>3

E <

E-fll

♦♦♦♦♦

siS ,®

co B cN®r‘

S < 1

lo

LT3CD

LOLO CDLO LO LO

COCDCO

LOCM

( S W V a O ) 1 H 9 I 3 M

JUN

JUL

AUG

SEPT

OC

T NO

V DE

C JA

N FE

B MA

R AP

R MA

Y JU

N JU

L AU

G SE

PT

30

Figure 10. Average female body weights(numbers in parentheses arenumber of animals).

Cl.

CDLO

CN

CQ

CN < LU

CO

c=>CD CD CDLO

CDCD

CDCD

f n C O LO L O 'T CO CO CVJ

(swvap) 1HOI3M

9Z6L SHiNOW

L

31

decrease in body weight from early summer to fall, 1975, followed by a

weight increase in December, 1975. The weight data for the remaining

months of winter and early spring 1976 indicate an overall decline in

adult male body weights. In June 1976, average weights of males began

to rise again. Subadults of the 1975 year class demonstrated a slight

weight gain from July to October followed by a winter weight decline

to 410g in December and January. Subadult weights increased from

February, 1976, onward as squirrels began to fully mature. Juvenile

squirrels exhibited a dramatic weight increase from June (263g) to

December (405g), 1975, and from June (313g) to September (440g), 1976,

following expected developmental trends of growth (Barkalow and

Shorten 1973).

Figure 10 illustrates the total body weights for captured females.

Average adult female weight declined from an initial high of 603g in

June, 1975, to a low of 440g in January, 1976. From January, 1976, to

September adult female weights increased to 555g. Captures of subadult

females were not frequent enough to show definite seasonal trends in

body weight. Similarly, data from June to August, 1975, is only

sufficient to indicate a steady weight gain for juvenile females in

the summer, and insufficient to show developmental trends for the

remainder of the sampling period.

Spatial Distribution

Home range statistics were computed using locational data obtained

by live-trapping and 294 hours of field observation of gray squirrel

behavior. A minimum of five locational fixes were used as input for

Rohlf's program to calculate a 95% confidence ellipse (Sokal and Rohlf

1969) to estimate individual home range size and activity center

32

(Koeppl, e_t aH. 1975). By analyzing the data seasonally, i.e. spring,

summer, and fall/winter, 39 individual home ranges were plotted for 24

resident animals (Figures 11-15).

The average areas for the 95% probability ellipses are listed in

Table 4. Home range area was computed manually by methods described in

Koeppl, _et_ aH. (1975) . The average yearly home range size for adult

males was 2.34 ha. Subadult males had slightly larger average home

ranges of 2.50 ha, and juvenile males had the smallest home ranges,

averaging 2.05 ha. Male animals had an overall combined home range

size of 2.28 ha. Size for the home ranges of females was 1.92 ha.

Adult females had average home ranges of 1.60 ha, while subadults had

an average of 3.12 ha. The single juvenile female for which data was

sufficient to compute a home range utilized an estimated area of 1.81

h a .

A three-way ANOVA was calculated for the home range data to detect

any significant variation in distribution of areas for males and

females by sex, age, or season (Table 4). There were no significant

differences in size of home range due to sex (p = 0.209), or season

(p = 0.846). Additionally, no significant effects were detected

between age (p = 0.390) and size of squirrel home ranges.

The proportion of an animal's home range ellipse that is over­

lapped by one or more other animal's home ranges was estimated from the

home range plots (Figures 11-15) by dividing each ellipse into quadrats

and approximating the percentage of area overlapped by other ellipses,

and is presented in Table 5. A three-way ANOVA revealed that male

home ranges overlapped less than did female home ranges (77.3% versus

92.3%), p = 0.025. Significant seasonal variation in overlap was

33

Figure 11. Home range 95% probability ellipses for resident animals during spring 1975.

• ••

34

Figure 12. Home range 95% probability ellipses for resident animals during summer1975.

?

??

??

I-------- 124m

35

Figure 13. Home range 95% probability ellipses forresident animals during fall/winter 1975.

36

Figure 14. Home range 95% probability ellipses for resident animals during spring 1976.

37,

S o

37

Figure 15. Home range 95% probability ellipses for resident animals during summer1976.

???

*4

38

<-4 CO > x Co CO > H3C c cp P p p «P p P< cr1 p 3 < cr P M crP P M P p p M P MP CP rt M P CP rt cn pH- P CD P H- p cnM I- 1 Cfi M M -O(D rt CD rt •Cn Cfi cn cn

o p PdX | X | X | X | X | X I X | X I Hi p o

/-s x—s x-—s /■—\ s—s p 33 | + P j + w l + p + + P j + p j + P + P cn P'•—' 'w ' P o

co co CO CO CO CO co CO H* P HM w M W M H W M 3 P p

p M pM CTQcn rt Ps ' H

I-* M to to to to P P• • • • • • €! XI H■o M CD co 4>- H- XI PCO -P> M OC -0- VO CO rt M P

✓—s M ,—s /—s x-~n XI P cnM ♦ 1 1 co| + + CO | + ro | + -M + + l-{ CTQ

00 '_' N—✓ N_' w H* rt oI—* o o o o o o p pr P Hi

• • • • • • CtQ H PCO CO 00 CO O'. -P* P CP HM M cr to M o P PCn ON M co co M 1 O cn

c; cr H-p cn CPX P P

H PCO I-* M M to to t£> < rt• • • • • • po -l> OO M to o o rt PCo VO 00 CO CT •o CO < H* P

,— X—s P > O H-1 ^ 1 + '■^l 3" — + ~ l + 1 1 M | + M + 3 P 3

tO OV 00 to O to 3 pw o ^ o '_/ o ^ o o o fD XI M

. • • • • • H p cnto 4> .£> to CO CO Hcn Ov to CO co to CO H-o CO -o o Co M p O o

p CP pcn w rto PP /—\ CtQ

to to X cn p O• • P H HOO M P H-CO cr M P N

x- n co /--s co ,— «•—\ M x- s — P1 M * 1 1 M • fo | + M • 1 1 CO + 23 H- CP

tso —’ to '— ' CT H- Pcr av o O o P CT

• • rt cr X-o (T tD pCO M r{ o cnC/1 CO rt p

P XHP Pcn P

CO M M to to to to > s* CP. • • • • • • MM Ov co o Co Co to M P PtO O to CO o -P- 00 ao

1— 1 ^^ z—^ ,— s X—s CO II p,,—N . ^s| + M | + M + ' - ' I + M + to + n>1—1 oo co M Cn -o CO -P> -O' p P a .^ M w o w o v_x o W o ^ o ^ o '— o cn c P• * • • • • • O § H

M ro to -P~ CT to to P cr H-o-v VO Ov co VO vo to P p PCT Ln CO 00 VO 'O •o H 00

39

H O rt 3 I—1

pel a CO3 3 O3 H* 3H- > CO CO 3 Ha CTO 3 3 O3 3 3 X 3 33 3 Hia O Hi O

3 3 HiOrt <!3 3

U>

CO33

OO OO to O I— 1 -P' o• • • • • • HiCn oo CO -P* CO ONO Cn O CO to COO Cn co Cn ►3

33H33

O COtO Cn a.Hi

a a1 a o a O• • • • • •I-1 to a to CO CO'O a "O O CO toOn Cn 'O Cn Cn

if33co►33333

O o a o• • • •vD a ON -oON ON -P- ONCO oo a a

Hj

COH-003H-Hi

o o o o a• • • • oCO 00 to Cn 3CD -P- o OO 3O ON CO -P~ o

OHiHj

33a3003

Table 4.

(continued) Analysis

of variance

of home

range areas

by sex,

seasons,

40

detected (p = 0.034) as well as a difference between age classes (p =

0.010). The overlap of spring home ranges was significantly less than

the overlap of summer ranges (comparison of means by Duncan’s range

test, p < 0.05). There was no significant difference between the

overlap of home ranges for either the fall/winter and spring, or

fall/winter and summer seasons. Females did not show a significant

difference in overlap between spring and summer home ranges; however,

males appeared to contribute more to seasonal variation in overlap than

did females (p = 0.025). Fall/winter female home range data was

insufficient for further testing of overlap.

To investigate age related differences in home range overlap a

Duncan’s multiple range test at the 0.05 level was also performed

(Table 5). The range test revealed no significant difference between

juvenile and adult home range overlap. Subadult home ranges, however,

showed significantly less overlap in comparison to both juvenile and

adult home ranges.

The percentage of overlap of home ranges (Table 5) was adjusted

for seasonal changes in the number of resident animals to produce the

home range overlap index data (Table 6). The overlap index was

calculated by dividing each percentage overlap value in the raw data

set by the number of animals categorized as residents during a

trapping and observation season. The density compensation of home

range overlap (Table 5) resulted in a qualitative change for three

aspects of the home range overlap index (Table 6) when tested for

differences by sex and season, as well as age. First, the overall

yearly difference in home range overlap between sexes decreased in

significance from the level of p - 0.025 to p = 0.044 (three-way

Juveniles X

+ SE

60.0 -

- 60.0

41

CO > Ch CO > Kp a. P a cc fDcr e 3 < cr C M& i—1 fU cd P t-1 CDa. rt I-* a a. rt CDP CO CD H- P CDh~> CO i-* i—■rt CD rtCD CD CD

X | X| xl XI X| X | X|✓—s x—*sw l + ^ | + + 3 1 + 3 1 + s 1 + +>_/ 'w' w

CO CO CO CO CO CO COK M w H Cd M M

VD CO -P- CN CnCn VD O o --J to. . • . • •O to o o Cn hO CO/-s /-\ TJ

1 1 cn | + av + CD| + NO | + 4H + + l-{'— ' V-/ w H-f—1 H* M PCn Cn O CD 00 03

O h-1 NO O ■J -P~O NO OO O OO CN

I—1OO vO VO o oo VO

Cn CD o 00 oCn CD — 1 o o o CO/—V rv /—n ✓-s /—s ~v ehO j + av| + col + n d | + 1 1 H-1! + H* +v— ' '—' o NJl-> '—' CDNO s> -O o Cn Cn Pm * . • • •Cn (-* o o VO oO VO o CD av

VD rrjCn I-1 fD. • t->I-1 h-1 o -J H*r-v O /-s o /-s tO /—^ —^I-* o 1 1 M o n d | + 1— 1 Cn 1 1 CD + s:-' m N»/ • -' m H*o o o to 3Cn CD rt. • CD

o Cn Ho t-1

VO VO VO 00 CD CO -o >H-1 Cn NJ Cn Cn ND -j 1— 1• . • . • • « I-*Cn CD '-J O I—1 CD/—v /—\ r\ CO/”NI + M| + ►—1 ■+• '-'1 + ^ 1 + i 114" (O + CDco 1—* Cn CD -O -P- fD

CD CD CD ' 00 v_x CN Cn CD« • • . . • • OCD o Cn I-1 Ov o CD CDCD Ui NO -P~ 4S OO CD

Table 5.

Average percent

overlap of

squirrel home

ranges, categorized

by sex

and age

during seasonal

trapping and

observation periods

(n =

number of

animals) with

three-way ANOVA,

42

t-3 K COO P P ort cn pP H* > 00 CO P pH* P- CTO p P o

P P p X! P pP cn HiCO O M o Hi o

p C P P HiUi a 4 P ncn P o rt <. a. p CD po p p p

h-1 H-rt cn PrtP H-P OP P0QfD

c-iP rt

oo < PNO P CD C/3• P rt to H* Pl_n H* O M Co NO to 00 3

I-* P Co Cn 00 '-J OfD rt co -O VO O o VO oOv M Cn Cn to -p- Hi

'P * • • • • •"O 00 CO OO Cn vo CO

11 VO to oo '-J VO Ov ►aCO 00 Cn o •£- Cn PO POo > . POO o p• p Cn CDo H* .

rt

O3 Co CO cco NO NO NO I— 1 Cn HioofDPH-O

CO P 31to n3 CD H* H 1 NO I-1 p• P Cn CO VO CO O -0 po H* CD cn Cn o Cn po P P vo o -o NO to OO

0Q O4 • • • • . . C/3cn av ON CN VO Cn VD P3p -p~ OO vo CO vo vO Prt co to CO Cn CO PCD PPO

Pi HiPH* 31— 1 p00 p Cn CO Cn -O• s: p • • • •H- cn 4>~ Cn OO PjCn P O Cn Cnrt P o H* to -ofD PP P

PPCp COp H*p eraVO CO H 1 P1— 1 P H- H-• 3 P HiCn 3 P o O O o H-

O fD CL • • . . OP . o O o o Pco NO o PO 4>~ Cn to P>5- * a- P

ohi

Table 5.

(continued) Analysis

of variance

of percent

overlap by

sex, season,

and age

Asterisk indicates

significant probability

less than

critical level

of 0.05.

43

CO t> ►3 cn > 33 3 Cu fD 3 3 o- 3< CT* 3 3 < cr 3 MfD CD h-1 CD fD CD H* fD3 Co rt h-1 3 D-* rt cnH- 3 cn fD H- 3 cnM I—1 cn I—1 (—1fD rt fD rtcn cn cn cn

X! | X| X 1 X I X I X| Xl X/“S — \ / v3j + 3 j + 3 J + 3 + 3 j + 3 j + 3 J + 3 , +cn cn cn cn cn cn cn cnM M m M K

H-* 1-*no NO oo On VO oooo NO oo M -X NO cn

/—V oo y—s /—^ s—N *3i— 1 • 1 1 VJll + cr| + OO I + NO | + ^ 1 + vo| + 3

cr V-/ '— ✓ V-/ H-o o i—1 H* NO o 3• • • • • • 00cr vo vO NO o vo

o NO On vo

Moo vO vO O VO VOoo cr JO' O o NO cn

/— s x-N s—s, 3NO | + CT) + oo| + NO | + 1 1 M + M | + 3'—' '—' o NO 3M o o o w o • o 3• • • • • • >-{NO JN -(O' o cr UlUl NO I— 1 o I-1

fD3cnO3

NO M •3OO 3• • 1—1

NO oo vO I-1✓-s Ul /-n NO /—s. i— cr /— s '—1—1 • I 1 I— 1 Ul N0| + M • I 1 Oo| + s:

o v/ • 00 — ' H-o t-* On 3• • rt

NO 00 3Ul cr 3

i—1 i— 1 I—1 1—1 >JN I— 1 f— 1 OO On VO VO f—1• • • • • • • MNO i— 1 On JO' On I-1 VO— 1 +

< v /— %— 1 +

/— \ cn/— s oo M | + M + M | + NO | + 3M • OO M On '-J oo JN JN 3w cr On O w i—* NO w o ^ o h-1 cn

• • • . • • • OJ>~ cr M -vj 00 cr o 3cr oo O VO oo 00 NO cn

Table 6.

Average overlap

indices of

squirrel home

ranges, categorized

by sex

and age

during seasonal

trapping and

observation periods

(n =

number of

animals) with

three-way ANOVA.

44

i-3Ort3

CO a3 3

cr> cr 3• 3 O'- j Cl. 3oo 3 3

I- 1 —rt cn

H33093

vo >• CL. rtvo 3 300 H* cn

rt rt

3rt cnco

TO CO.

II h-13 Co

M < o NONO 3 •. 3 o00 H- Ul

(—1 •fD

35O3 COo -'-J

0933H-O

CO 3VO 3 cn• gto § cn H*Ul fD e

H cr •3 I - 13 1-*rt NO3

OCO Hi

VO no• H 3oo H- 3o 3 3

09 33

3hi3

33 3M Cl.

I-* H1 3vo '•>s >i• H*H- H-3 3rt 3fD Cl,H •

*3 ES CO3 3 o3 H- 3H- > CO CO 3 HCL. 09 3 3 D3 3 3 X 3 33 3 HiM O Hi O

3 3 Hinrt <3 3

HH-3rtH-O3

CO33

NO M NO CO oNO NO co Ui Hi00 -P- OO Ul. . • . . C/0

vo OO NO I-1 >n"J Ui CO Cn Ui 3•P' o '-J VO 00 3

H33

co CL.NO NO NO I— 1 Ul Hi

K3I-1 3CN H 1 CO "J 3

00 -O' NO 00 M• • . • • COCN NO -P- NO o ►aOO NO H 1 Ul CO 3"J Ul OO VO NO 3

H3

H*NO -P- OO• • .CO VD ■P> I—1 HJvo -O' oCO I-* -p- '-J

COH*

093H*Hi

o O o o H-• . • . OO o o o 3o o •P'- o 3NO o 4>- o OOf- X* OS­ OS- 3

OH i

Table 6.

(continued) Analysis

of variance

of overlap

index by

sex, season,

and age.

Asterisk indicates

significant probability

less than

critical value

of 0.05.

44

ANOVA's). Second, the significant effects due to age increased noticably

from p = 0.010 to p = 0.002. Third, the degree of home range overlap in

the fall/winter season was significantly different from the overlap in

spring and summer (Duncan's range test, p < 0.05). The overlap index

results indicate that squirrels in the fall/winter had home ranges that

tended to overlap more with the home ranges of neighboring animals than

during either the spring or summer seasons. Meanwhile, subadult

animals had overlap indices more similar to adults than to juveniles

(Duncan’s range test, p < 0.05).

The average distance between activity centers of the nearest same

sexed animals was calculated from the 95% probability ellipses using 2 g hthe formula Dc = U [ (X-x) + (Y-y) ] 2 given by Koeppl et_ al. (1975) for

computing the Euclidean distance between the activity center and a speci­

fied location. The yearly average distance between males of 16.5 m

(Table 7) was not significantly different (p = 0.712) from the average

distance between females of 17.8 m. Likewise, no significant difference

was indicated for seasonal variation in distances between activity

centers for males and females (p = 0.199). A significant effect was

revealed between age and distance to the same sex (p = 0.020). Subadult

activity centers were less closely spaced than were the activity centers

of adults and juveniles (Duncan's range test, p < 0.05). In summary,

the subadult age class had a significantly greater mean distance to

same sexed animals than adults or juveniles; thus, subadult squirrels

of.the same sex were spaced significantly farther apart on the yearly

average than were juvenile or adult animals of the same sex.

The distance between activity centers of the nearest opposite sexed

animals was computed utilizing the 95% probability ellipses and the

45

c—I CO > CO > K3 3 Cb CD 3 3 Cb 3< cr* 3 3 < cr* 3 h-1CD 3 I-1 3 3 3 I—* 33 O, rt I-1 3 Cb rt 3H* 3 CD CD H* 3 3h-1 1—1 CD 1—* I—1CD rt 3 rtCD CD 3 3

xl X| X 1 xl X| X 1 x l X3J + 3J + 3J + 3^ + w 1 + 3 1 + 3j + 3, +'_✓

C/D CO CO CO CO CO co COw m w M

H* I-1 -F- i—* roCO FO H-* I— 1 ■o o• • • • • •ON CO OO Ln co oo CO

x-s CF /—^ /-% z-\ /-s TJI-1 • 1 1 Wi | + Ov| + co | + ro| + ^ 1 + ^ 1 + t-4CO '-- —/ '—^ '— ' '—/ H-

M CO VD Cn Cn 3• • • • • • OQvo o vo O h-1 rooj h-* ■vj 00 O oo

to to to 1— 1 1— 1CO I-1 to • h-1 O• • • O • •t-1 Ov O Cn -F~ to CO/—s 3

^ 1 + ~v| + ^ 1 + ^ | + 1 i i-1! + M +ro OV to to o tooo o-' Cn w O w ro w ro 3• • • • • •co h-1 -F- o Cn to-> t-1 -F» Ov •o

CDfDCDO3

to to00 oo 3• • i—1i— 1 vo h-1

/—s CO ✓—sI - 1 o ro | + 1 I co | + s:'---' • v—/ H-

Ov to h-1 3• • rtCn '-O 3CTv i—1 rf

to h-1 1— 1 I— 1 CO h-1 1— 1 >CO ol oj -F- O] CO crv M• • • • • • • ‘I— 1 VD 00 ro vD h-* Cn

✓'■'v /> CO^ CTv ~ l + M | + M + ^ 1 + — 1 + M + to| + 3h-> • to t-1 -F~ "-j co -F' -F~ 3w 0O w 00 w CTs ro to 3• • • • • • • O

CO o CO 00 CO CO 3I—1 VD vD ON -F~ CO 3

Table 7.

Average distance

(meters) between

activity centers

(A.C.) of

same sex

animals, categorized

by sex

and age

during seasonal

trapping and

observation periods

(n =

number of

animals) with

three-way ANOVA.

46

h-*oofO-P'

c-t3<fD3H*

O33o33

333CTQ<D

rt(DCDrt

3rttO

OUl

s cnO 3 3 ort CO H* 33 H- > cn cn 3 t-iI-4 CO 00 3 3 O3 3 3 X 3 33 CD t-hh-* o t-h O3 3 t-hOrt <CO 3

On -P~ I-* I-1vo I-1 (—1 -P-ro OV On -O t—1 On-P~ VO CT\ ro OO -P-• • • • • •o o O o o vo4>- Ov o o VD oo

OO t—1 On OO ro

3H-3rtH*O3

cn33ot-hcn33333CD

Ot-h

OUl

I-1Ultoro

VDUl

!>P-»3

cn3cr3CO3 I—1 rt

og0933H-O3CDCD

&CO3

Ot-h3333CO3i-i333Cl.3t-SH-1H*33Co

OO■J O Oro to ro On

O o O o• • . .o h-1 ■ 1 oto vD H 1 OOoX-

VD ro ooX-

Co

s3H* I-1 Ul ro oo 3Ov OO -o ro 1— 1 On 3O o OO OO o. • • • * • cnh-1 ro o o o VD ►ao oo o o VD VD 3VO oo o ro 00 cr 3

tt3

■p- H-> o ro-P> CS\ r-* Ovoo VD oo VDOV Ov VD -P-

cnH*003H*t-hH-O33o3OH i

Table 7.

(continued) Analysis

of variance

of distance

to same

sex A.C.

by sex,

season and

age. Asterisk

indicates significant

probability less

than critical

level

47

general formula cited above (Koeppl ej al . 1975) for distance to same

sexed animals. Table 8 presents the resulting distance to opposite

sexed animals activity centers. No seasonal, age, or sex related

differences were statistically significant for distances between

activity centers of opposite sexed animals as determined by a three-way

ANOVA (Table 8).

Another aspect of the population spatial distribution that was

investigated was the change in location of individual activity centers

over time. Table 9 lists average distances for shifts in activity

centers by sex and season. The overall average movement of an activity

center was 18.9 m. Males shifted activity centers an average distance

of 23.7 + 9.21 m (Mean +_ SE), while females had an average shift of

14.7 + 5.00 m. However, the difference in degree of shift of activity

centers exhibited between the sexes was not significant (p > 0.20). No

significant seasonal movement of individual centers of activity was

indicated from overall Mann-Whitney U tests, p > 0.20 (Table 9, fall/

winter data not included because of small sample size).

Removal and Release

The results of the removal and release procedure at the conclusion

of 16 days trapping and observation indicated that resident squirrels

could return to their home ranges when released at a distance of 488 m

(1600 ft). Exactly 50% (3/6) of the resident squirrels returned to

their home range areas. Animal 11, an adult male, was recaptured on

the study area one day after being released. Animal 45, also an adult

male, was observed on the study area the second day after release and

subsequently recaptured the following day. Animal 14, an adult female,

was recaptured on the study area on the eighth day after release. None

48

C-I CO > X f-l CO > s HP P CL fD p p CL p P< c r P 3 < c r p I - 1 crfD P (-> P (D P M fD H*P Cl rt i—1 P Cl rt cn pH- P cn fD H* P cnI - 1 H* cn M H* 00fD rt fD rt •CO CO cn CO

o PX I X 1 X 1 X | X | X | X l X l cr P <

/~s /-S ✓—v /—x cn X P

d l + P J + P J + 0 + P j + P j + PJ + P , + (D PP P PCO CO CO CO CO co CO CO < P OQw M M M W M w M P H- P

rt 3H- P CLO H-* H*P P P

I- 1 I - 1 t—* -P' f—1 to w rth-1 I - 1 Ul ■O 00 to X3 P• • • • • • P O PUi VO Ui I - 1 Ul to CO P P O

✓~X H* ,—X ^-v /—x / x /--X /-X xt H- r t PI - 1 CO 1 1 U i| + ON + co | + to | 4- + vo + p O P' —' • w v--✓ '—' N---' H* CL OQ Z'--X

VD CO p W O 3I- 1 1—* VO to Ui 'U CTQ P P• • • • • • —x H- r t

-O- CO I- 1 Ui Ui p N Po -O o Ui -P- -P» P P

II CL Px_/

p crp X cr

M h-1 h-* h-1 I—1 a pCO I- 1 I - 1 to to o cr p r t• • • • • • p p ZOn CO OO o CO P x P

r v /—x p P1 ^ 1 + + /-X + ^ \ + 1 1 M + 1—1 "+" o p P

ro ON 00 to o to t-h pn- ' oo w ov '_' -P' h-1 ^ CO "—' to p CL P

• • • • • • p P oo o ON CO o 00 P P r tCO I—1 -o to -o o CO H> OQ H-

fD 3 P <Jp P H-p t-1 CL rto w P Xp P

to to trj H- O-O to P t , P P• • t-» H- OQ P-P' Ui 1—1 rt r t

/->. I - 1 S--X h-1 ^x ^ I - 1 S--X "—. pr P P1 h-1 to 1 1 1--1 to to | + 1—1 00 i I CO + s: P P

s_ • V—/ • '—* • H- rt P Po o t-1 ~o P pr P

to rt p O ✓—x• • fD p P >-P~ -P- P p P •to to I I - 1 O

•p r t X--'X PP o

I - 1 h-1 h-1 h-1 CO i—1 h-> > X5 HiCO t—1 t—1 -P- Ul -P- ON h-1 X X)• • • • • • • I—* o H- P1—1 -P~ VO to On to CO < P P

t -1 —X ,—x ,--X /—x CO > CT9 P~N CO ^ 1 + M + i - 1!+ — 1 + ---1 + M + to + fD • PM • CO M Ui -o CO -P- -P' P P Pw vo '—' '—' '—* V_X fxj s—' '—' cn P P

CO to Ui o to CO o CL rt• • • • • • Pon to -P" ON -O -o -P- cn Oas to Ui o o o X3X)o

pH-rtfD

49

H & s COO fD p Ort CD H* CP H* > CO CO P Hh-1 CL CTO fD (D O

C fD P X fD fDP CD HiI—1 O Hi O

P P H inrt <CD fD

HH*PrtH-OP

CO

MOO CO Cn ro Cn O"■o o -P- 1—‘ ro Ov H iCO 00 VO oo ro• • • • • . COI—1 i—1 00 oo -o o ►Qco NO o CO Cn o CCO ON OO ■o P

HfDCD

CO CO CL'O tO ro ro h-1 Cn Hi

ro I - 1 CO ro ro CO

Kpp

i—1 vo Cn ro I - 1 pro -o vO oo ro. . . • . . CO

1—* -P" vo ■o 4>~ ►QCo CT> o I - 1 Cn O CCO O NO 00 00 H* P

HP

(-* H* I—1 h *• • • • HIoo CO Cnvo H 1 Ov 00vo 00 O Cn

COH*H*Hio o O o

• . • •H1 ro ro MOv 00 oo voav ro vo CO

fD

OH i

Table 8.

(continued) Analysis

of variance

of distance

to opposite

sex A.C.

by sex,

season, and

age.

50

Table 9. Comparison of average shift (meters) in gray squirrel activity centers. Mann-Whitney U tests for difference in distribution by season and sex not significant, p>0.20 (season: nl = 10, n2 = 5, U = 37; sex: nl = 8, n2 = 7, U = 37). Sample size in parentheses.

SeasonMales X + SE (n)

Females X + SE (n)

Spring to Summer 14.9 + 3.79 (4) 14.0 + 6.27 (6)

Summer to Spring 35.4 + 21.10 (3) 16.8 + 10.80 (2)

Total 23.7 + 9.21 (7) 14.7 + 5.00 (8)

51

of the alien Williamsburg squirrels immigrated to the study area. Two

out of the seven aliens were observed on separate occasions within 80 m

of the release point. Resident animals were not observed in the

release point area. A Fisher’s exact test for independence of homing

rates by resident and alien squirrels yielded a borderline significance

value of p = 0.0699. Thus, while resident animals displayed a tendency

to return to the study area at a rate better than the alien squirrels

could randomly immigrate, this difference was not significant.

Behavioral Observations

Field observations of gray squirrel behavior for 294 hours yielded

247 recorded agonistic interactions between 18 animals. The frequency

of observable interactions for the fall/winter period was insufficient

to include for analysis; consequently, comparisons were only made for

the spring and summer behavioral data. A Dominance Index, similar to

the A/S ratio used by Swenson (1977) for prairie deermice, was developed

to quantify the intensity of agonistic interactions exhibited by

individual squirrels. The index was computed using square root trans­

formed raw data in the following manner. The Dominance Index (D.I.) is

equal to the number of dominant outcomes divided by the number of sub-/VnDom. + ^missive outcomes of observed agonistic interactions: D.I. =.,-------- -/nS ub. + \

Figure 16 illustrates the mean Dominance Index (D.I.) scores and

their 95% confidence intervals for males and females during 1975 and

19 76. There appears to be a trend for males to have higher D.I. scores

than females upon inspection of the amount of overlap shown by the 95%

confidence limits (Fig. 16). The D.I. score of the sole pregnant adult

female for summer 1975 is an exception to the trend. Mann-Whitney U

52

Figure 16. Dominance Index scores for males and females (95% confidence limits shown, n=number of animals).

DOMI

NANC

E IN

DEX

4 - r

S P R I N G J SUMMER

1975S P R I N G S U M M E R

1976

CM

53

tests (Table 10) of D.I. by sex and season indicated a significant

difference between the D.I. scores for males and females. Males had

significantly higher D.I. than females in the spring (nq = 7, np = 6,

U = 39, 0.01 < p < 0.02), but no significant difference was found for

the summer scores (ni = 9, n2 = 3, U = 18, p > 0.20).

Previous studies (Bronson 1964, Brenner et aA. 1978) have shown

that the dominant animal in an agonistic encounter often has a higher

body weight than the subordinate animal. Table 11 lists some Spearman

rank correlations of body weight and D.I. scores. Male squirrels had

a significant positive correlation rs = 0.453, p < 0.05 between D.I.

score and body weight, while females had a positive, but not

significant correlation rs = 0.539, p > 0.10 for the same two para­

meters .

Dominance Index scores were tested for correlation with home range

overlap and area (Table 11) to ascertain if some aspect of spatial

relationships were associated with dominance. Home range area was not

significantly correlated with D.I. scores for either males or females.

A significant negative correlation (rg = -0.496, p < 0.009) was found

between D.I. scores and home range overlap. Dominant animals, as

defined by having a high D.I. score, would tend to have a smaller per­

centage of their home range overlapped by the ranges of other animals.

To quantify levels of aggressive behavior, Figure 17 presents the

mean number of agonistic interactions per hour of observation for 1975

and 1976. The addition of a feeding station to the study area in

June, 1976, corresponded with an increase in the number of observed

interactions between squirrels. However, levels of agonistic inter­

actions were not significantly different between summers (Mann-Whitney

54

Table 10. Comparisons of dominance index (D.I.) by sex and season.Mann-Whitney U tests for differences in distribution significant by sex, 0.01 < p < 0.02 (nl = 9, n2 = 16,U = 115.5), but not significant by season, p > 0.20 (nl = 12, n2 = 13, U = 98.5).

Male Female Overall D.I.Season X + SE (n) X + SE (n) X + SE (n)

1975 Spring 2.6 + 0.80 (4) 0 . 7 + 0 . 4 2 (4) 1.6 + 0.41 (8)

Summer 1.1 + 0.20 (3) 1.7 (1) 1.3 + 0.17 (4)

1976 Spring 1.7 + 1.16 (3) 0 . 8 + 0 . 3 9 (2) 1.3 + 0.43 (5)

Summer 1.4 + 0.89 (6) 0.4 + 0.07 (2) 1.1 + 0.31 (8)

Overall D.I. 1.7 + 0.94 (16) 0.8 + 0.49 (9) 1.4 + 0.18 (25)

55

Table 11.

Sex

Males

Females

Combined

Spearman rank correlations (rs) of Dominance Index with body weight, percent overlap, and home range area. Asterisks indicate significance.

Index with Overlap

n = 13

rs = '0.391

0.05 < p < 0.10

n = 9

rs = 0.272

0.40 < p < 0.50

n = 22

Index with Body Weight

n = 15

rs = 0.453

0.01 < p < 0.05*

n = 8

rs = 0.539

0.50 > p > 0.10

n = 23

Index with Area

n = 13

rs = 0.234

0.40 < p < 0.50

n = 8

rs = 0.539

0.05 < p < 0.10

n = 21

rs = "* 0.496

p < 0.009*

rs =0.281 rs = 0.298

0.05 < p < 0.10 0.05 < p < 0.10

56

Figure 17. Levels of aggressive interactions during summer1976 and 1976.

MEA

N IN

TER

AC

TIO

NS

PE

R HO

UR

OF

OB

SE

RV

AT

ION

.4

1975.2 1976

.0

.8

.6

.4

.2

.0SEPTA U GJ ULJ U N

MO N T H S

57

U = 0.5, p > 0.20). Adjusting the data (Figure 18) for population

density during the two summers, dividing number of interactions per

hour by the MLE population values, reveals more of a disparity in the

levels of aggressive behavior (Mann-Whitney U = 12.5, p < 0.10). If

the months of June, July, and August only are considered for comparison,

then a significant difference (Mann-Whitney U = 9.0, p < 0.05) is

apparent in levels of aggressive interaction between the summers of

1975 and 1976.

The orientation of aggressive interactions during the summer of

1975 and 1976 is presented in Figure 19. The format of the orientation

of interactions as given in Figure 19 (js.£., Resident: Immigrant)

places the dominant animal to the left of the colon and the subordinate

animal to the right of the colon designating a type of dyadic inter­

action. Resident: Immigrant (i . e_. , dominant animal: subordinate

animal) agonistic interactions declined from 38% to 13%, comparing

July, 19 75, with July, 1976, while the reciprocal Immigrant: Resident

interactions rose from 12% to 35% of total interactions. By August,

1976, feeding priorities at the feeding station between residents and

immigrants were more established than in July, 19 76, with Resident:

Immigrant interactions dropping from 13% to 9%, and Immigrant: Resident

interactions declining from 35% to 18% of observed agonistic encounters.

As Immigrant: Resident and the corresponding reciprocal type of

encounter (Resident: Immigrant) declined, Resident: Resident inter-

actions increased from 17% in July, 1976, to 55% in August, 1976.

58

Figure 18. Levels of aggressive interactions adjustedfor population density during 1975 and 1976.

MEA

N

INT

ER

AC

TIO

NS

PE

R A

NIM

AL

0.14 - -

0 . 1 2 - -

0 .10 —

0.08 - -

0.06

0.04 - -

0.02 - -

0.00 - -

J UN

1975 S

1976

JUL A U G

MONTHSSEPT

59

Figure 19. Orientation of aggressive interactions duringsummer of 1975 and 1976.

aasafisafiga

(%) SNOIJ.DVil3J.NI 3AISS3HOOV

DISCUSSION

The study population of gray squirrels on Jamestown Island was

comparable to other populations in the eastern United States. The

characteristic May-July breeding season (Taylor 1966; Nixon and McClain

19 75; Thompson 1977a) was evident, but no winter breeding period was

observed. Failure of female winter breeding has been reported for

other gray squirrel populations (Cordes 1965; Barkalow and Shorten

1973). This decline in reproductive rate is often associated with a

poor mast crop the preceding autumn; typically, male squirrels develop

normally but the females do not come into estrus (Cordes 1965; Barkalow

and Shorten 1973). Both perforate females and scrotal males were

captured on the study area during the winter. Since the fall 1975 mast

crop on the study area was not quantified, one can only speculate as

to why no winter breeding occurred.

The decline in numbers of the population from an estimated 38

animals in June, 19 75, to 15 animals in June, 19 76, could be due to

the winter decline in female reproductive rate, and either increased

rates of emigration or mortality. Trap derived census techniques such

as the Maximum Likelihood Estimate (Nixon €it_ _al. 1967) cannot

distinguish between a reduction in captures due to variables such as

sex, age, capture status, and climatological factors, or increased

emigration and mortality. Perry ejt aTL. (1977) noted the high intrinsic

variability that characterized gray squirrel trap response, and cite

the need to adjust for environmental variation to obtain accurate mark-

recapture population statistics.- Nevertheless, the MLE provides an

60

61

adequate estimate for gray squirrel population size in comparison to

other commonly used population statistics (Nixon e_t aTL. 1967) .

Johnson (19 73) notes that scent marking behavior can serve a

number of functions, and that it is probable that scent marking serves

different functions in different species. Scent marking by the gray

squirrel has been observed to act as a stimulus to attract males to

estrus females (Barkalow and Shorten 1973). Yet, the function of

scent marking must be more than a sexual attractant since Taylor (1968)

observed that peaks of marking activity occurred both during and out­

side of the breeding seasons. It is possible that scent marking in the

gray squirrel functions more broadly as an indicator of individual

identity, perhaps including information on sexual status as well as

age and dominance as suggested by Johnson (19 73).

In this study, no clearly defined scent marking behavior was seen

during 294 hours of observation. The results of the urine application

experiment did not indicate a significant approach or avoidance

response of gray squirrels to enter urine treated live-traps. Perhaps

if the proven stimulus of urine from females in estrus had been used

the phenomenon of male attraction may have been observed as reported

by Taylor (1966, 1968) and Barkalow and Shorten (1973). More experi­

mental field research is required to clarify the function of scent

marking in the eastern gray squirrel.

The size of gray squirrel home range ellipses in this study

(Table 4) fall within the range of areas reported by other researchers

utilizing convex polygons as estimates (Flyger 1960; Taylor 1966;

Doebel 1967; Donohoe and Beal 19 72; Cordes and Barkalow 19 72; Bland

1977; Thompson 1978). Although not directly equivalent, convex

62

polygons yield comparable size estimates to ellipses when corrected

for sample size bias (Jennrich and Turner 1969). Reported average home

range values range from 0.77 ha (Flyger 1960) to 9.75 ha (Bakken 1959)

for adult males, and from 0.48 ha (Flyger 1960) to 4.01 ha (Bakken

1959) for adult females. Subadult home ranges vary from 15.2 ha (Adams

1976) to 1.09 ha (Cordes and Barkalow 1972). The data indicate that

males on the average have larger home ranges (2.28 ha) than females

(1.92 ha), but the difference is not significant. No clear age

related difference in home range size is indicated by the data.

However, the data tend to agree with the results of Cordes and Barkalow

(19 72) and Adams (19 76) in that subadult squirrels were more mobile and

had larger home range areas than adults (2.5 and 3.12 ha, versus 2.34

and 1.6 ha for both males and females). Seasonal changes in home range

size as reported by Bland (1977) and Thompson (1978) were not demon­

strated in this study.

Significant differences in spatial distribution of the study popu­

lation were indicated by changes in the percentage of home range over­

lap and overlap index. These differences can be attributed to both

sex related and seasonal effects. Male home ranges overlapped

significantly less than female home ranges; this difference lies in

the seasonal patterns of male overlap, whereby male ranges overlap

significantly less in spring and summer, while overlap is significantly

greater in fall/winter. The trends for decreased male home range

overlap would be consistent with the demonstrated avoidance of more

aggressive animals by subordinates during spring-summer breeding

period (Farentinos 1972; Thompson 1977a) and increased overlap would

be consistent with subadult movements during fall dispersal (Cordes

and Barkalow 1972).

63

The significant age related difference in distance to the activity

center of nearest same sexed animals was linked with the increased

distances between the activity centers of subadult males and the

activity centers of other males in the spring and fall/winter seasons.

More frequent mobility of male squirrels has been reported during

breeding seasons (Farentinos 1972; Bland 1977; Thompson 1977a, 1978)

and fall dispersal (Cordes and Barkalow 1972). Avoidance by subadult

males of the more aggressive and dominant adults during these periods

could account for the significant age related difference found in the

distances between activity centers of same sexed animals.

The results of Cordes and Barkalow (1972) and Thompson (1978)

indicate that established gray squirrels have relatively stable home

ranges over time. In this study, the activity centers of 15 animals

for which data was complete, 7 males and 8 females, showed no

significant seasonal shifting of individual activity centers. The

data indicate that established animals in the Jamestown study popu­

lation had stable home ranges for the duration of this study.

The stability of the gray squirrel home range system is further

demonstrated by the results of the removal/release procedure. Half of

the resident Jamestown animals successfully returned from a distance

of 488 m, a rate of return indicating better orientation than shown

by the random movements of the alien Williamsburg squirrels, and an

indication of a definite tendency for residents to return to their

center of familiarity. Hungerford and Wilder (1941) first observed

the homing ability of gray squirrels by trapping animals and releasing

them at points 914 m to 6.1 km distant. Forty per cent of the animals

were later retrapped; the maximum distance from which animals returned

64

was 4.5 km (Hungerford and Wilder 1941). Taylor et_ jil. (1971) found

that of 57 squirrels trapped within their study area and released 2.6 km

away, 16 squirrels were known to have settled within 3.2 km of their

point of release and 4 of these squirrels returned to sites near their

point of capture. Brady (19 72) trapped and displaced 53 gray squirrels

from their home woodlot to other woodlots up to 2.2 km distant. Forty

three per cent of the animals became residents where released; seventeen

per cent of the animals actually returned to their home woodlot, and no

animals returned home when displaced at distances greater than 1.1 km

(Brady 1972). The removal/release procedure results of this study con­

firm that gray squirrels are likely to home (p = 0.069) when released

within 1 km of their home woodlot.The work of Pack et jlL. (1967) and Thompson (1978) reveal that

gray squirrels have a social system based on a dominance heirarchy

among a population of neighboring aniamls. Social dominance is related

to both sex and age of the animal (Pack et al. 1967; Thompson 1978),

with oldest adult males being at the top of the heirarchy followed by

similarly aged adult females. Adult males in this study tended to

have higher Dominance Index scores than females, but this difference

was significant only for the spring observational periods. A larger

sample size in this study would have increased the power of the tests

used to determine if more conclusive sex related differences in

dominance existed as demonstrated by other researchers (Bronson 1964;

Pack et aT. 1967, Thompson 1978). Body weight of the animal was

positively correlated with Dominance Index scores. Dominance was also

found to be negatively correlated with home range overlap. Animals

with high D.I. scores had home ranges that overlapped less than the

home ranges of neighboring low D.I. score animals. The data suggest

65

that subordinate gray squirrels were avoiding contact with dominant con-

specifics in the study area. Dominance was not associated with actual

size of home range as in Pack et al. (1967).

Bronson (1964) noted that the prevalence of wounding in female

woodchucks reached a peak during lactation from May-June and that

aggressive interaction rates declined steadily thereafter. Similarly,

the reproductive condition of female gray squirrels can influence their

Dominance Index score. The D.I. score of animal 99, an adult female,

tripled when the animal was pregnant during the summer 1975 season.

Taylor (1966) and Sharp (1959) also report increased aggressiveness in

pregnant female gray squirrels. It would be of interest to determine

if male squirrels in breeding condition also exhibited similar changes

in dominance. However, due to the uncertainty in accurate assessment

of actual male reproductive condition (active spermatogenesis versus

regressing or recrudescing testes) in the field by external genital

examination (Pudney 1976) it is inappropriate to include such an

analysis in this study.

Utilization of a feeding station for a portion of this study did

increase the number of observable agonistic interactions among squirrels

on the study area. Pack et al. (1967) also report the high number of

social interactions observed among squirrels at feeding stations. This

study differs from that of Pack et al. (1967) in that use of a feeding

station was limited to summer 1976 (3 months) instead of the entire

study. Population density adjusted data indicated a significant

difference in levels of aggressive behavior for the summer of 1975 and

1976. A qualitative difference also existed in the orientation of

agonistic interactions between the summer 1975 and 1976 observation

66

periods. This is exemplified by a reversal in percentages of Resident:

Immigrant and Immigrant: Resident type of dyadic interactions for the

two summers. The feeding station appeared to attract immigrant

squirrels from adjacent woodlots as well as to present an interactive

focal point for the resident squirrels on the study area. Pack et al.

(1967) conducted their study in discrete woodlots and therefore did

not observe immigrant animals interacting with residents as in this

study.

Thompson (1978) observed peaks of aggressive behavior comparable

to this study in his gray squirrel population during summer and early

fall. The home range expansion of spring born animals was associated

with the summer peak, while the early fall aggression peak corresponded

to the fall movement period during which immigrants arrived on the

study area (Thompson 1978). During both peaks of aggression, adult

resident animals were dominant in 86.7% and 91.8% of the Adult:

Juvenile, and Resident: Immigrant interactions (Thompson 1978).

Orientation of aggressive interactions changed from Resident: Immi­

grant to Immigrant: Resident simultaneously with establishment of the

feeding station on the study area. Two aggressive adult male immi­

grants actually succeeded in establishing priority at the feeding

station over resident squirrels. The feeding station was such an

attractive stimulus to immigrants that the resulting intensive

agonistic interactions with residents deviated from the norms of the

previous summer, and the observations of Thompson (1978).

The findings of this study lend direct support to the work done

by Bland (1977) and Thompson (1978) on the spatial distribution of a

gray squirrel population. The results that dominance was not shown

67

to be associated with home range size as by Pack eh ad. (1967), and

that residents were not consistently dominant over immigrants as shown

by Thompson (1978) may be attributable to use of a feeding station in

this study, since noticeable increases in the levels of aggressive

interactions began upon its establishment on the study area. Utilization

of the home range model of Koeppl er al . (1975) resulted in values for

gray squirrel home ranges within the limits of areas reported in the

literature and provided a quantifiable basis for analyzing the spatial

distribution of the population.

APPENDIX

Ethogram of Gray Squirrel Behavior

Aggressive (A)

(1) Low intensity threat - squirrel in upright position, ears fullyraised, eyes may be slitted; accompanied by squeak growl and possibly rapid tail flicking.

(2) High intensity threat - squirrel crouched, eyes slitted and earspartially raised; accompanied by growl and/or teeth chattering.

(3) Chase - squirrel in close pursuit of opponent; often accompaniedby attempts to bite tail of fleeing squirrel.

(4) Attack - squirrel engaged in locked scratching and biting fightwith opponent.

Submissive (S)

(1) Pawing - squirrel faces attacker and strikes with forepaws; maybe accompanied by squeak growl and teeth chattering.

(2) Defensive - crouched position, tail curved over back, eyes wideopen; Kuk-kuk....quaa vocalization may also be given.

(3) No change - ignores other animal.

(4) Avoidance - alters established position by moving, or makes adetour around other squirrel; movements stiff and jerky, hair erect.

(5) Flight - flees, pursuer, may be accompanied by squealing.

Sexual (X)

(1) Investigative approach - male sniffs ground and nearby treesadjacent to female, and approaches her position with slow hesitant movements.

(2) Mating chase - male follows female persistently, sniffing atfemales genital region as both animals move.

(3) Circling - male circles female slowly, continuing to sniff thefemales anogenital region, accompanied by some licking also.

68

69

(4) Mounting - male mounts female for period of 5-16 seconds; femalebeing held in position by the hindquarters.

(5) Genital grooming - male licks penis; generally occurs afterseveral mounts have culminated in ejaculation.

Grooming (G)

(1) Automanipulative - squirrel grooms itself.

(2) Social - squirrel grooms another animal.

Miscellaneous (M)

(1) Foraging - animal moves over ground, sniffs at substrate, and digsfor food object.

(2) Travel - animal rapidly traverses ground w/o stopping.

LITERATURE CITED

Adams, C. E. 19 76. Measurement and characteristics of fox squirrel,Sciurus niger rufiventer, home ranges. Am. Midland Nat. 95(1):211- 215.

Bakken, A. 1959. Behavior of gray squirrels. Proceedings 13th Annual Conference Southeastern Association of Game and Fish Commissioners, pp. 393-407.

Barkalow, F. S. and M. Shorten. 1973. The world of the gray squirrel. J. B. Lippincott Company. Philadelphia. 154 pp.

Barrier, M. J. and F. S. Barkalow. 1967. A rapid technique for aging gray squirrels in winter pelage. J. Wildl. Mgmt. 31(4):715-719.

Bland, M. E. A. 1977. Daily and seasonal activity patterns in the eastern gray squirrel. Unpubl. Ph.D. Dissertation. Univ. of Minnesota. 227 pp.

Brady, J. T. 1972. The behavior of displaced gray squirrels. Unpubl. M.S. Thesis. Virginia Polytechnic Institute and State Univ.

Brenner, F. J., C. P. Gaetano, S. W. Mauser, and D. L. Belowich. 1978. Body weight and social interactions as factors in determining dominance in captive eastern chipmunks Tamias striatus. Anim.Beh. 26(2):432-437.

Bronson, F. H. 1964. Agonistic behavior in woodchucks. Anim. Beh. 12:470-478.

Burt, W. H. 1943. Territoriality and home range concepts as applied to mammals. J. Mamm. 24:246-352.

Calhoun, J. B. and J. U. Casby. 1958. Calculation of home range and density of small mammals. U. S. Public Health Service Monogram 55. 24 pp.

Cordes, C. L. 1965. Home range and movements of the eastern graysquirrel, Sciurus carolinensis carolinensis Gmelin, in Wake Co., N.C. M.S. Thesis, North Carolina State Univ. 58 pp.

Cordes, C. L. and F. S. Barkalow. 1972. Home range and dispersal in a North Carolina gray squirrel population. Proceedings 26th Annual Conference Southeastern Association of Game and Fish Commissioners, pp. 124-135.

Cotter, J. L. 1958. Archeological excavations at Jamestown, Virginia. Archeological Research Series Number Four. National Park Service, U.S. Dept, of Inter.

70

71

Doebel, J. H. 1967. Home range and activity of the gray squirrel,Sciurus carolinensis in a southwest Virginia woodlot. M.S. Thesis, Virginia Polytechnic Institute and State Univ.

Doebel, J. H. and B. S. McGinnes. 1974. Home range and activity of agray squirrel population. J. Wildl. Mgmt. 38(4):860-867.

Donohoe, R. W. and R. 0. Beal. 1972. Squirrel behavior determined by radiotelemetry. Ohio Fish and Wildl. Rep. 2:1-20.

Farentinos, R. C. 1972. Social dominance and mating activity in the tassel-eared squirrel, Sciurus aberti ferrens. Anim. Beh. 20: 316-326.

Flyger, V. F. 1960. Movements and home range of the gray squirrel, Sciurus carolinensis, in two Maryland woodlots. Ecol. 41(2): 365-369.

Hayne, D. W. 1949. Calculation of size of home range. J. Mamm, 30:1-18.

Horwich, R. H. 1968. The ontogeny of social behavior in the graysquirrel. Unpubl. Ph.D. Dissertation. Univ. of Maryland. 295pp.

Hungerford, K. E. and N. G. Wilder. 1941. Observation on the homing behavior of the gray squirrel. J. Wildl. Mgmt. 5:458-460.

Jennrich, R. I. and F. B. Turner. 1969. Measurement of non-circular home range. J. Theoret. Biol. 22:227-237.

Johnson, R. P. 1973. Scent marking in mammals. Anim. Beh. 21:521- 535.

Koeppl, J. W. , N. A. Slade and R. S. Hoffmann. 1975. A bivariate home range model with possible application to ethological data analysis. J. Mamm. 56(1):81-90.

Koeppl, J. W . , N. A. Slade, K. S. Harris and R. S. Hoffmann. 197 7.A three-dimensional home range model. J. Mamm. 58(2):213-220.

Mazurkiewicz, M. 1969. Elliptical modification of the home range pattern. Bulletin de L'Academie Polonaise des Sciences 17(7): 427-431.

Mazurkiewicz, M. 19 71. Shape, size, and distribution of home ranges of Clethrionomys glareolus. Acta Theriol. 16:23-60.

McBride, G. 1971. Theories of animal spacing: the role of flight,fight, and social distance. In: Behavior and Environment: theuse of space by animals and men. Ed. A. H. Esser. Plenum Press. New York. pp. 53-68.

72

Mosby, H. S. 1969. Livetrapping objectionable animals. Publication 283. Cooperative Extension Service. Virginia Polytechnic Inst. Blacksburg, Virginia.

Nie, N. H., C. H. Hull, J. G. Jenkins, K. Stein-Brenner and D. H. Bent.1975. Statistical Package for the Social Sciences, 2nd. ed. McGraw-Hill Book Co., New York. 675pp.

Nie, N. H. and C. H. Hull. 19 77. SPSS Batch 7.0 Update Manual.Unpubl. 135pp.

Nixon, C. M. , W. R. Edwards and L. Eberhardt. 1967. Estimating squirrel abundance from live-trapping data. J. Wildl. Mgmt. 31:96-101.

Nixon, C. M. and M. W. McClain. 19 75. Breeding seasons and facundity of female gray squirrels in Ohio. J. Wildl. Mgmt. 39(2):426-438.

Pack, J. C., H. S. Mosby and P. B. Siegel. 1967. Influence of social hierarchy on gray squirrel behavior. J. Wildl. Mgmt. 31(4):720- 728.

Perry, H. R. , G. B. Pardue, F. S. Barkalow and R. J. Monroe. 1977.Factors affecting trap responses of the gray squirrel. J. Wildl. Mgmt. 41(1):135-143.

Pudney, J. 19 76. Seasonal changes in the testis and epididymis ofthe American gray squirrel, Sciurus carolinensis. J. Zool. Lond. 179:107-120.

Sharp, W. M. 1958. Aging gray squirrels by use of tail pelage characteristics. J. Wildl. Mgmt. 22(1):29-34.

Sharp, W. M. 1959. A commentary on the behavior of free-running gray squirrels. Proceeding 13th Annual Conference Southeastern Assoc. Game and Fish Commissioners. pp. 382-387.

Siegel, S. 1956. Nonparametric statistics for the behavioral sciences.McGraw-Hill Book Co., New York. 312 pp.

Sokal, R. R. and F. J. Rohlf. 1969. Biometry. W. H. Freeman and Co.,San Francisco, California. 776 pp.

Swenson, B. 1977. Social and reproductive consequences of bridging asymtotic laboratory populations of prairie deermice. Unpbl.M.A. Thesis, College of William and Mary.

Taber, D. and McT. I. Cowan. 19 71. Capturing and marking wild animals. In: Wildlife Management Techniques 3rd. ed. Ed. by R. Giles.Edwards Bros., Inc., Ann Arbor, Michigan. pp. 277-318.

Taylor, J. C. 1966. Home range and agonistic behavior in the gray squirrel. Symp. Zool. Soc: Lond. 18:229-235.

Taylor, J. C. 1968. The use of marking points by gray squirrels. JZool. Lond. 155(2):246-247.

Taylor, K. D., M. Shorten, H. G. Lloyd and F. A. Courtier. 1971.Movements of the gray squirrel as revealed by trapping. J. ApplEcol. 8(1):123-146.

Thompson, D. C. 1977a. Reproductive behavior of the gray squirrel. Can. J. Zool. 55:1176-1184.

Thompson, D. C. 1977b. Diurnal and seasonal activity of the gray squirrel. Can. J. Zool. 55:1185-1189.

Thompson, D. C. 1978. The social system of the gray squirrel. Behaviour. 64(3-4):305-328.

VanWinkle, W. 1975. Comparison of several probabilistic home rangemodels. J. Wildl. Mgmt. 39(1):118-123.

Watson, A. and R. Moss. 1970. Dominance, spacing behavior, andaggression in relation to population limitation in vertebrates. In: Animal populations in relation to their food resources.Ed. A. Watson. Blackwell Sci. Pub. Oxford. pp. 167-222.

VITA

Marston Earl Youngblood, Jr.

Born in Washington, D.C., November 10, 1952. Graduated from Terry

Sanford Senior High School in Fayetteville, North Carolina, June 1970,

and received AB in Biology from the University of North Carolina at

Chapel Hill, May, 1974. In September 1974 the author entered the

College of William and Mary as a graduate student in Biology. Course

requirements and research for the MA degree were completed in September

1976.

The author has been employed at the Virginia Institute of Marine

Science of the College of William and Mary since October, 1976, first

as a statistician in the Department of Invertebrate Ecology, and

presently as a laboratory technician in the Department of Crustaceology.

74